Observability Boilerplate for Log-Metric-Tracing Collection

This is a simple observability boilerplate that is aimed at Log-Metric-Tracing collection

Components

1. Application: Python Django
2. Log Collection: Elastic
3. Metric Collection: Prometheus
4. Trace Collection: Zipkin
5. Dashboarding: Kibana | Grafana | Zipkin

Note: A huge thanks to @mordaha for building a base seed project on top of which this boilerplate is built.

It uses:

• python 3
• django 1.9
• postgresql for django database (dockerized for development, host-based for production)
• nginx as production webserver, with send_file && secure_link for serve uploaded files
• elasticsearch + logstash + kibana (ELK) for log collection
• prometheus for metric collection
• zipkin for trace collection
• ./setup.sh for "docker-compose with development config" shortcut

Boilerplate Setup and URLs

1. Run $ ./setup.sh
2. Wait until all containers start, then run ./es_index.sh
3. Python Django application URL: http://127.0.0.1:8000/
4. Nginx URL (First log entry): http://127.0.0.1:8080/
5. Logstash Endpoint: http://127.0.0.1:8050/
6. Elasticsearch URL: http://127.0.0.1:8020/
7. Kibana Dashboard URL: http://127.0.0.1:8056/
8. Prometheus Dashboard URL: http://127.0.0.1:9090/
9. Grafana Dashboard URL: http://127.0.0.1:3000/
10. Zipkin Dashboard URL: http://127.0.0.1:9411/

License

MIT

TODO

• Polishing the setup